I’ve been struggling for past few months to compute bayes-optimal predictions for non-trivial (analytically solvable) latent (i.e. prior) distributions. I feel like I made progress, but it’s not yet correct. (I did see this post, but didn’t seem it was on the same topic, or I couldn’t understand it:…

Hi @ricardoV94 , thanks so much for the help. Indeed if I use this mixture model as you suggested w = pm.Dirichlet('w', a=np.array([1, 1])) prior = pm.Mixture('prior', w=w, comp_dists=[pm.Beta.dist(alpha=alpha1, beta=beta1), …

Modeling multivariate distributions for bayesian optimal inference

Questions v3

ricardoV94 June 19, 2023, 7:49am 4

I don’t see your new question

Topic		Replies	Views
Pymc3 gives shifted result compared to analytical bayes solution v3	4	407	June 19, 2023
Struggling with Beta mixture models Questions	3	1874	February 2, 2019
Using Bayesian for estimating parameter Questions	7	3475	January 21, 2019
Modeling multivariate distributions - pymc does poorly on highly peaked latents v3 modeling	0	327	July 3, 2023
How to apply bayesian to find posterior of multivariable? Questions	1	452	April 9, 2019

Modeling multivariate distributions for bayesian optimal inference

Related topics